Levosimendan for Weaning From Veno-Arterial Extracorporeal Membrane Oxygenation in Cardiogenic Shock: An Umbrella Review.

Veno-arterial extracorporeal membrane oxygenation (VA-ECMO) is a key rescue therapy in cardiogenic shock (CS), but weaning failure remains common and is associated with worse outcomes. Levosimendan, a calcium-sensitizing inotrope with vasodilatory and cardioprotective properties, has been proposed to facilitate weaning, and several systematic reviews and meta-analyses with overlapping primary studies have examined this question, with varying conclusions regarding the strength and consistency of the evidence. This umbrella review was conducted to consolidate that evidence base. Following PRISMA 2020 and PRIOR guidelines, PubMed, Embase, Scopus, and Cochrane Library were searched from inception to May 2026 for systematic reviews and meta-analyses evaluating levosimendan as an adjunct for VA-ECMO weaning in adult CS patients. Two reviewers independently screened studies, extracted data, and assessed methodological quality using AMSTAR 2; given substantial overlap in constituent primary studies, a narrative synthesis was performed rather than a de novo meta-analysis. Seven reviews published between 2020 and 2025, pooling 5 to 18 primary studies and 557 to 2,274 patients, met inclusion criteria. All seven reviews found levosimendan significantly improved successful VA-ECMO weaning versus control (OR/RR range 1.20-2.89), with heterogeneity varying across reviews (I² 0-77%), and most also reported a significant reduction in mortality. Secondary outcomes were less consistent, with two reviews reporting numerically longer VA-ECMO support duration or ICU stay in levosimendan-treated patients. AMSTAR 2 confidence ratings ranged from low (n=2) to moderate (n=4) to high (n=1), substantial primary-study overlap was identified across reviews, and GRADE certainty, where formally assessed, was rated very low. Levosimendan may be associated with improved VA-ECMO weaning success and a possible mortality benefit in CS, particularly in patients with combined cardiac insufficiency, but this concordance rests on overlapping, retrospective, non-randomized cohorts, making it hypothesis-generating rather than an evidence-based standard pending adequately powered randomized controlled trials.